feat: Add speculative decoding notebook for 2-3x speedup

[Karanjot786]

Description

This PR introduces a new notebook, [Gemma_3]Speculative_Decoding.ipynb, demonstrating how to achieve significant inference speedups (1.5x - 3x) for Gemma models using speculative decoding with Hugging Face Transformers.

Closes #291.

Key Features

• Speculative Decoding: Implements assisted generation with a draft model.
• Prompt Lookup Decoding: Includes self-speculation for T4/low-memory environments.
• Hardware Aware: Automatically detects GPU type (T4 vs A100/L4) and selects compatible models/precision (Gemma 2 or Gemma 3) to prevent errors.
• Benchmarking: Provides built-in latency and throughput comparisons.

Reviewers

@bebechien - Please review the notebook approach and benchmarking logic.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[gemini-code-assist]

Summary of Changes

Hello @Karanjot786, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request introduces a new Jupyter notebook designed to significantly enhance the inference speed of Gemma models. By leveraging speculative decoding techniques through Hugging Face Transformers, the notebook demonstrates how to achieve substantial performance improvements. It also incorporates hardware-aware logic to adapt to different GPU environments and includes benchmarking tools to quantify the speedup, making it a valuable resource for optimizing Gemma model deployments.

Highlights

• New Notebook Introduction: A new notebook, [Gemma_3]Speculative_Decoding.ipynb, has been added to demonstrate speculative decoding for Gemma models.
• Inference Speedup: The notebook showcases how to achieve significant inference speedups (1.5x - 3x) for Gemma models using speculative decoding with Hugging Face Transformers.
• Decoding Techniques: It implements assisted generation with a draft model and includes Prompt Lookup Decoding for self-speculation in T4/low-memory environments.
• Hardware Awareness: The solution automatically detects GPU type (T4 vs A100/L4) and selects compatible models/precision (Gemma 2 or Gemma 3) to prevent errors.
• Benchmarking: The notebook provides built-in latency and throughput comparisons to evaluate the performance gains.
• Documentation Update: The Gemma/README.md file has been updated to include an entry for the new speculative decoding notebook.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

The pull request successfully introduces a new notebook on speculative decoding, enhancing the capabilities of Gemma models. The update to the README.md correctly lists the new notebook. My feedback focuses on improving the conciseness of the notebook's description in the README table for better readability.

[bebechien]

LGTM!

[bebechien]

I'll fix nbfmt later.